Abstract Colonization of artificial seagrass was examined at four sites in a New Jersey estuary, two within natural eelgrass Zostera marina L. beds and two on sand flats distant from natural eelgrass. Near-weekly sampling of fishes and decapod crustaceans demonstrated rapid and dense colonization, with densities often exceeding those in natural eelgrass. Distance from natural eelgrass had a pronounced effect on the composition of the resultant community, with significant differences in densities near and far from natural eelgrass. Although the structure of seagrass communities has often been largely attributed to the structural complexity of the seagrass canopy, the contrasts among artificial seagrass at the four sites in this study indicated that canopy structure was secondary to other factors. Colonists of artificial seagrass included large numbers of settled juvenile and adult fishes and decapods, in contrast to Bell's et al. (1985, 1987, 1988) conclusion that colonists of artificial seagrass were primarily planktonic settlers. At the sites distant from natural eelgrass, colonists apparently emigrated from alternative habitats, crossing expanses of predationrisky sand to reach the small isolated plots of artificial seagrass. This emigration contrasts with Bell & Westoby's (1986a) hypothesis that organisms should remain in the habitat into which they initially settle, even if poor in quality, rather than risk movement across unvegetated substratum.